Why is there distortion on some of my mixes?

[HeyHeyMyMy]

Hi folks,

I'm having a little problem. I recorded a friend of mine just playing acoustic guitar and singing. When we laid the tracks I was careful to watch the meters and kept everything between -6 and -3db. Then I tried out a copy of Wavelab 5 that my friend has and used a little L2 on it. On the meters in Wavelab the levels never indicated that they got above -3db. I am monitoring with a pair of Yorkville YSM1p monitors and everything sounds fine and dandy on them. However, when I take the song and listen to it on other systems there is some digital clipping. Any thoughts on why this might be occuring? Any suggestions would be much appreciated.

Thanks!

 

[SouthSIDE Glen]

However, when I take the song and listen to it on other systems...

My first question would be what steps are you taking between Wavelab and the other system. Are you transferring the WAV files to data CD? Or burning music CD? Or maybe creating an MP3?

It's possible there's something going on in that stage that's translating your final WAV file in Wavelab improperly.

G.

 

[danny.guitar]

What systems did you try it out on?

Some small speakers can distort pretty easily. For example, my friend's computer has small speakers and even at low volumes they can distort and rattle.

I would rip the track from the CD onto your computer, and open it into a wave editor and check the levels.

 

[HeyHeyMyMy]

SouthSIDE Glen - You've probably hit the nail on the head here. In Wavelab I believe I had it set to export to MP3 (because he wanted to post the song online). Is this likely to have been the problem? If so, how should I go about doing things so that I can give him an MP3 copy of the song?

Danny.guitar - Another excellent point. I've listened to it on three different systems, and at least one of them was a computer with crappy speakers. At the same time though, I listened to a couple of other tunes on those same speakers at higher volumes and they didn't distort. Any idea why that would be? (I'm sure there's a simple answer, and in asking this question I've probably illustrated my relative technological ignorance, ha ha...)

Thanks for the help thus far, guys!

 

[jonnyc]

I hate giving this generic of an answer but if it's distorting turn it down. It doesn't need to be as hot as you can make it, it just needs to be at a reasonable enough level to hear. Turn it down!!!

 

[SouthSIDE Glen]

In Wavelab I believe I had it set to export to MP3 (because he wanted to post the song online). Is this likely to have been the problem?

I don't know that I can say that it's "likely", but it's a possibility to be looked at.

I've not used Wavelab's MP3 encoder, so I can't comment specifically, but Wavelab is a good product form a good company; I'd imagine they are using a quality enocder (probably and hopefully a licensed Fraunhoffer codec), so I wouldn't think that'd be a probelm in and of itself.

However the followup question would be how does that MP3 sound on your Wavelab system? Does it sound OK there? Also try re-importing that MP3 into Wavelab and take a look at the waveform of the exported MP3 in there. How does it look?

As far as the details beyond that, maybe someone more familiar with current Wavelab could chime in. But in general, I'd encode at 192k minimum (128k only if bandwidth were an issue) using a good encoder.

Also, there is the question of what you are using to play it back on the playback systems. If you're using something like Winamp or Windows Media Player, etc. Make dure the EQ and effects settings are all flat, bypassed, or turned off, and that the soundcard volume levels are set correctly. The clipping or distrotion could be exascerbated on that end too.

G.

 

[HeyHeyMyMy]

Thanks a lot, Glen. I'll take a look at those things.

jonnyc - you make a good point, and in most cases you'd probably be right. But in this case (and this is the part that was most annoying to me when listening to the track) the volume was only at about half. I definitely wasn't pushing it at all, which piqued me all the more.

 

[xstatic]

If you tracked between -6 and -3 it is quite possible that you overdrove your preamp. Especially if you also used a compressor when tracking.

 

[Massive Master]

"Quite possible" is an understatement.

 

[gtrman_66]

If you tracked between -6 and -3 it is quite possible that you overdrove your preamp. Especially if you also used a compressor when tracking.

I agree, that's tracking way too hot. Peaking in a range of -15 to about -9 is much more like some sane levels, especially at 24 bits. No need to push it that hard. Someone swat me if I'm wrong, but I beleive most digital stuff is calibrated th where 0vu comes in at about -15 db on the digital meter, at least that's where mine lands. If you push it to -3, that would equal +12 vu

needless to say, things can get a bit "crispy" running those kinds of levels.

And the equipment manufacturers aren't helping things. I my HD-24 manual under setting levels it says "get the level as close to zero as possible without going over". Bullshit, there is just no call for running that hot that I can see.

 

[HeyHeyMyMy]

I agree, that's tracking way too hot. Peaking in a range of -15 to about -9 is much more like some sane levels, especially at 24 bits. No need to push it that hard. Someone swat me if I'm wrong, but I beleive most digital stuff is calibrated th where 0vu comes in at about -15 db on the digital meter, at least that's where mine lands. If you push it to -3, that would equal +12 vu

needless to say, things can get a bit "crispy" running those kinds of levels.

And the equipment manufacturers aren't helping things. I my HD-24 manual under setting levels it says "get the level as close to zero as possible without going over". Bullshit, there is just no call for running that hot that I can see.

Ahh! Well, that sure would make a lot of sense. This is was the first time I've actually tried to track at that level (thanks to a "tip" I read on this board) so it would make sense that I'm encountering this problem for the first time. It was stupid of me not to realize that this was likely the problem. Thanks for the verbal slap upside the head, guys!

 

[xstatic]

Most of my equipment seems to use -18 dbfs as the equivalent to 0dbvu. Most analog circuits now are rated to at least +22, but that does not mean that they are efficient at +18 for example. I would not say that from a technical standpoint that Alesis was wrong to say that you want the meters as close to 0 as possible. From a converters point of view that is where they are the most efficient, closer to 0. The problem is that the converters most efficient point is well above that of almost everyones front end that feeds it. Luckily, converters seem to have a much narrower difference in quality form say -3 to -18 dbfs as a preamp does between the same level (0dbvu and +15 dbvu). If on the other hand everyones analog front end had a nominal level equivalent to -6dbfs, than we would all want to track as hot as possible on the converters and get the best of both worlds.

 

[SouthSIDE Glen]

And the equipment manufacturers aren't helping things. I my HD-24 manual under setting levels it says "get the level as close to zero as possible without going over".

That's seems to be what most hardware manuals on the planet say. I think they are all just copying some old analog hardware manual from the 1960s and no one has bothered to update it.

The problem is the phrase "as close to zero" means absolutely nothing unless or until one defines zero *what?* If they don't specify 0dBVU - which is about the only scale in which that statement would make any sense - they might as well be talking 0°C or zero feet above sea level. But it most certainly does not mean 0dBFS in digital software like Wavelab.

G.

 

[gtrman_66]

Well, I was paraphrasing last night when I posted, this is what the manual actually says.

Have the performer play the loudest section of the song you are about to record.Adjust the gain controls of your mixer (trim,channel and master) until the loudest notes fall just below the clip indicator on the HD-24's meter.

Then there is a footnote about performers playing louder on the actual take and setting levels conservatively to compensate blah blah, but all in all they are indicating that the target level should be right up to the clip line on the digital metering.

Too hot, too hot, too hot.

Hell, most of my snare drum tracks hardest hits are around -10. every once in a while one will pop up to -6 or so, but not very often.

 

[ecktronic]

SouthSIDE Glen - You've probably hit the nail on the head here. In Wavelab I believe I had it set to export to MP3 (because he wanted to post the song online). Is this likely to have been the problem? If so, how should I go about doing things so that I can give him an MP3 copy of the song?

Danny.guitar - Another excellent point. I've listened to it on three different systems, and at least one of them was a computer with crappy speakers. At the same time though, I listened to a couple of other tunes on those same speakers at higher volumes and they didn't distort. Any idea why that would be? (I'm sure there's a simple answer, and in asking this question I've probably illustrated my relative technological ignorance, ha ha...)

Thanks for the help thus far, guys!

I dont see why comverting to MP3 would produce digital distortion.
I would suggest getting a good pair of headphones and crankgin the volume to check for distortion. If theirs no distortion in the decent headphones then it must be your speakers that are feked.
If there is distortion then it could be that your pre amp is peaking even though you are going into it at -6 to -3dB. I have an old analogue desk thaa I use the pre amps to record with and I never go above -6dB as I notice disotrtion when I go above this.
Could just be dodgy old meters.

Eck